Defining the concept of pulse conditions..

Introduction.

Wearablepulse wave detection devices(witha single-probe pressure sensor) are more and more applied for healthmonitoring. The pulse signal contains very rich cardiovascularphysiological and pathological information. According to the pulsediagnosis theory of Traditional Chinese Medicine (TCM), using thepulse signal can not only detect whether the subject is abnormal butalso predict the pathological condition. Therefore, it is of greatresearch significance to make an objective description of pulsesignals.

Пn the study of the objectification of pulse diagnosis, pulse signal processing is a crucial step in obtaining the diagnosis result of Chinese medicine, including pulse signal segmentation and feature extraction and pulse signal pattern recognition and classification.

Accurate segmentation and feature extraction of pulse signals are very important for the objectification of pulse diagnosis.

From the earliest days, metaphors and descriptive languages have been used to describe the characteristics of various pulse conditions in China treatises. For example, a «Fu Mai» (“floating pulse”) is presented as being like wood floating on water, a «Kou Mai» (“hollow pulse”) is described as the feeling of pressing a scallion stalk, and a «Ge Mai» (“drumskin pulse”) is described as resembling the beating of a drum. These descriptions are incomplete and conceptually unclear, without specific data references.

Given the obscurity and incomprehensibility of pulse terminology, visual aids have been sought to define pulse conditions since ancient times. The earliest published book that uses drawings to illustrate the characteristics of pulse conditions is «Chabing Zhinan» (“A Guide to Disease Examination”), «Tuzhu Maijue Bianzhen» (“Pulse Discrimination by Figure”) published in the “Ming dynasty” and «Renyuan Maiying Guizhi Tushuo» (“Condensed Pulse illustration”) published in the Western Jin dynasty, also used drawings to illustrate the distinct characteristics of pulse conditions.

Drawings in these ancient books were simply dots and lines or a combination of pictograms in one circular area.

Examples of using figures and parameters to define pulse conditions.

According to the «Shuyu Gongzuo»: Yuanze Yu Fangfa (Terminology Work: Principles and Methods),2 non-verbal representations can illustrate and exemplify a concept. They should not replace a definition but complement it. Non-verbal representations include visual representations, such as figures, and mathematical expressions.

Modern science and technology provide tools for visual representation and quantification of pulse conditions that can help present precise definition of pulses.

In the 1950s, Chinese scholar Zhu Yan introduced a lever-type pulse descriptor to research TCM pulse conditions.4 Since then, most studies have focused on pulse measuring devices and analysis methods of pulse, and have formed a unified understanding on the typical pulse figures of common pulses such as the slippery pulse and the string-like pulse.

The slippery and string-like pulses.

A pulse waveform graph records the trajectories of vascular pulsation of the radial artery at the wrist and contains data regarding the cardiac ejection and the process which the pulse wave travels along the vascular system.3

Different physiological and pathological states produce pulse conditions with different waveforms. The typical pulse waveform graphs of the slippery pulse and string-like pulse which consensus has been reached are presented in Figure 1.

Figure1.Thepulse waveforms visualizing the characteristics of pulse conditionsthrough a two-dimension representation with X-axis and Y-axisrepresenting time and amplitude, respectively: (A)A slippery pulse with obvious double-peak waveform and a narrowpercussion wave; (B)A string-like pulse with a broad and high percussion wave

The pulse waveforms only support a qualitative understanding of pulse conditions, thus quantitative analysis of the waveforms is needed to obtain pulse parameters.

Methods for extraction and analysis of pulse parameters.

Zhou Xuehai, a physician in the late Qing dynasty, wrote that 5 “location, rate, shape, and force must be clarified to understand all pulse conditions.”

This classical quotation means that the identification of pulse conditions is mainly made through four aspects: location, rate, shape, and force. Pulse parameters should be extracted to reflect the four aspects of pulse conditions using different analytical methods. Of the four aspects, the pulse shape can be quantified by the time domain analysis method and the hemodynamic method.6-8
We, therefore, used these approaches to extract the parameters of the string-like pulse and the slippery pulse.

Thetime domain method ⁹is an intuitive method that mainly quantifies the characteristics ofthe pulse wave form in a single cardiac cycle (Figure2).

The pulse wave form parameters, including h2/h1, h3/h1, h4/h1, h5/h1, As/Ad, t1/t, t1/t4, and t5/t4, can reflect the morphology of different waveforms, and have different physiological significances (Table 1).

Figure2.Thepulse waveform graph in a single cardiac cycle with X-axis and Y-axis representingtime and amplitude, respectively.

h1: the height of percussion wave; h2: the height of canyon between percussion wave and tidal wave; h3: the height of tidal wave; h4: the height of dicrotic notch; h5: the height of dicrotic wave; t1: the time distance between the start point of pulse chart and percussion wave; t4: the time distance between the start point of pulse chart and dicrotic notch; t5: the time distance between dicrotic notch and the endpoint of pulse waveform; t: the time distance between the start point and the endpoint; w: the width of percussion wave in its 1/3 height position

Thepulse wave form parameters, including h2/h1,h3/h1,h4/h1,h5/h1,As/Ad,t1/t,t1/t4,and t5/t4,can reflect the morphology of different waveforms, and have differentphysiological significances (Table1).

Thepurpose of pulse analysis in the hemodynamicmethodis to analyze the factors influencing pulse waveforms. According tothe hemodynamic principle, pulse waves can be decomposed into forwardand reflected components. The pulse wave recorded in the radialartery is the synthesis of forward and reflected traveling waves.^9-11

Therefore,the pulse wave velocity (PWV)and the reflection coefficient (R_f),which represent properties of transmission and reflection of thepulse wave, respectively, are parameters that can be used to explainmorphological changes of the different pulse waveforms. The formationgraphs of a string-like pulse and a slippery pulse wave-forms arepresented in Figures3 and 4.¹¹

Figure3.Theformation graph of a string-like pulse waveform.¹¹

For example, in older individuals, arterial stiffening causes increased pulse wave velocity. Thus, the early return of the reflected wave affects the systolic than the diastolic part of the wave, augmenting the percussion wave with a secondary rise in late systole after an early systolic peak that creates the waveform of a string-like pulse.

Figure4.Theformation graph of a slippery pulse waveform.¹¹

Forexample, in youth, good arterial compliance causes decreased pulsewave velocity. Thus, the reflected wave affects the diastolic ratherthan the systolic part of the wave, causing secondary fluctuations indiastole, and forming the double-peak wave of a slippery pulse.

Figures3 and 4show that PWVand R_fare the factors affecting the formation of pulse waveforms, whichhelp interpret the waveform differences between a slippery pulse anda string-like pulse. Using the hemodynamic method for extractingphysiological information from TCM pulse conditions,¹²we obtained the values of PWVand R_fof the slippery pulses and the string-like pulses.

Complementing definitions of the two pulses.

Tocarry out the calculations, we used the time-domain and hemodynamicmethods to analyze 247 samples of slippery pulse and 622 samples ofstring-like pulse, and calculated the time-domain parameters (h2/h1,h3/h1,h4/h1,h5/h1,As/Ad,t1/t,t1/t4,t5/t4,and w/t),PWVand R_fof these pulses. SPSS 25.0 (IBM Corp, Armonk, NY) was used to analyzethe pulse parameters.

TheMann-Whitney U test for non-parametric method was applied to comparepulse parameters that are not normally distributed, and distributionsof pulse parameters are described by median, the highest and lowestquar tile, as M(Q_L,Q_H).The results are presented in Tables2-4.

Table 2

Comparisonof time-domain parameters of the two groups of the pulses M(Q_L,Q_H)

*representsa significant difference compared with a slippery pulse, P<0,001

Table 3

Comparisonof time-domain parameters of the two groups of the pulses M(Q_L,Q_H)

*representsa significant difference compared with a slippery pulse, P<0,001

Table 4

Comparisonof PWV and R_fof the two groups of the pulses M(Q_L,Q_H)

*representsa significant difference compared with a slippery pulse, P<0,001

In these experimental pulse samples, the values of time-domainparameters h2/h1,h3/h1,h4/h1,h5/h1,t1/t,t1/t4and w/tin the string-like pulses were higher than those in the slipperypulses (P<0.001), and the t5/t4value was lower than that in the slippery pulses (P<0.001). The PWV and R_f of string-like pulses were higher than those of slippery pulses(P<0.001). These differences between the two groups of pulses werestatistically significant.

The experimental results show that pulse parameters can distinguish thewaveforms of slippery pulses and string-like pulses.

The hemodynamic principle indicates that arterial stiffening and peripheral resistance can be reflected from the waveform of pulse,and arterial stiffening and peripheral resistance can becharacterized by PWVand R_f.

Ourstudy shows that the string-like pulse with higher PWV and R_fi ndicates high arterial tension, and corresponds to the descriptionthat “a string-like pulse is like the musical strings, and stiffunder the force of the fingers”.¹³

Theslippery pulse with a lower PWV and R_f represents good arterial compliance, and corresponds to thedescription ¹⁴“a slippery pulse arrives and departs smoothly. Its form is likepearls rolling on a plate.”

Theslippery pulse with a lower PWV and R_f represents good arterial compliance, and corresponds to thedescription ¹⁴“a slippery pulse arrives and departs smoothly. Its form is likepearls rolling on a plate.”

Therefore,the time-domain parameters, PWV,and R_f of the pulse conditions can quantify the waveform differences ofstring-like pulse and slippery pulse from the aspects of morphologyand formation mechanism of waveforms. As exemplified in the aboveexamples, these parameters are used to complement the definition oftwo types of pulses.

Сonclusions and further suggestions.

First,pulse measuring device with a single-probe pressure sensor is mostcommonly used in clinical settings to obtain pulse information anddisplay the dynamic pulse waveforms of a patient’s radial artery.Researchers have made progress in obtaining objective pulseconditions through extraction and analysis of pulse parameters. ^9,14

However, the information received from a device using a single-zone pressuresensor is far from exhaustive, reflecting four aspects of the pulsestate. Therefore, new measuring devices based on photoplethysmogramswere later developed, which, as expected, would collect much morecomplete information reflecting the “location, frequency, shape,strength” of pulse conditions.

Second, the positioning mode of current pulse measuring devices produced bysome companies still relies on manual positioning. The sensor must bemanually moved to the position of the radial artery and pressureneeds to be manually adjusted to obtain an optimal pulse waveform.These manual operations result in unreliable data, affecting theaccurate acquisition of pulse waveforms and parameters. Thus,automatic positioning and pressurization techniques are importantissues to be resolved in the future.

The traditional approach to pulse diagnosis (TCM) relies on the sensitivepalpation of a physician’s fingers, through which the physicianobtains 3D pulse information. New sensors and digital signalprocessing technology should be applied in pulse measuring devices toobtain 3D waveforms of pulse conditions.

Using 3D pulse waveforms could maximize the simulation of the pulsediagnosis process and obtain more reliable pulse information.

The development of sensor technology, artificial intelligence and bigdata may support the informatization and digitization of pulsediagnosis, ultimately ensuring that the theoretical system of pulsediagnosis can be improved for future generations.

In conclusion, standardized methods on digitalization and quantificationof pulse conditions based on AI can provide objective data forsubjective definition of pulses, thus improving the disease diagnosisand treatment, as well as teaching and training.

References.

1.Usanov D.A., Skripal A.V., Vagarin A.Yu., Rytik A.P.. Methods andequipment for diagnosing the state of the cardiovascular systemaccording to the characteristics of the pulse wave. – Saratov:Publishing House Saratov University, 2009. – 96 p.: ill. ISBN978-5-292-03959-4

2.National Technical Committee on Terminology and Language ContentResources of Standardization Administration (SAC/TC 62). TerminologyWork-principles and Methods. Beijing: China Quality and StandardsPublishing and Media Co; 2019. Chinese.

3.Fei ZF. Contemporary Sphygmology in Traditional Chinese Medicine.Beijing: People’s Health Press.2006. p. 56. Chinese.

4.Lu XZ, Hu GQ, Cao HM, et al. Research progress and prospect of pulseinstrument. Paper presented at: The 11th National Academic AnnualConference on Diagnosis of Traditional Chinese Medicine by ChinaAssociation of Chinese Medicine; July 1, 2010; Beijing, China.Chinese.

5.Zhou XH. Zhou Xuehai Medical Encyclopedia.Beijing: China TraditionalChinese Medicine Press; 1999. p. 631-653. Chinese.

6.Qiao LJ, Qi Z, Tu LP, et al. The association of radial artery pulsewave variables with the pulse wave velocity and echocardiographicparameters in hypertension. Evid Based Complement Alternat Med 2018;PMID: 30622604.

7.Cui J, Tu LP, Zhang JF, et al. Research on pulse graphcharacteristics of 1720 cases with different health status and agegradient. Shanghai Med J 2018;52(4):15–23. Chinese.

8.Nichols WW, O’Rourke MF. McDonald’s Blood Flow in Arteries.3rded. London: Edward Arnold; 1990.

9.Fei ZF. Research on Pulse Diagnosis in China Shanghai: ShanghaiCollege of Traditional Chinese Medicine Press; 1991. p. 100–166.Chinese.

10.Westerhof N, Sipkema P, van den Bos GC, Elzinga G. Forward andbackward waves in the arterial system. Cardiovasc Res 1972;6(6):648–656.

11.Guo R, Wang YQ, Yan HX, et al. Feature extraction and recognition oftraditional Chinese medicine pulse based on hemodynamic principles. JIntegr Med 2010; 8(8):742–746. Chinese.

12.Chinese Patent: 200710045229.7.

13.LiZZ. Essential Principles and Practice of Pulse Diagnosis. Beijing:China Press of Traditional Chinese Medicine Corporation;2008. p.79–96. Chinese.

14.YanHX, Wang YQ. Research and Clinical Application of Pulse Diagnosis inTraditional Chinese Medicine. Shanghai: Shanghai Scientific andTechnical Press.2016. p. 79–96. Chinese.

15.Bolshev A.S. Heart rate. Physiological and pedagogical aspects[Text]: textbook. manual / A.S. Bolshev, D.G. Sidorov, S.A.Ovchinnikov. Nizhny Novgorod State University of Architecture andCivil Engineering: – N.Novgorod: NNGASU, 2017. – 76 p. ISBN978-5-528-00188-3